THE QUESTION OF FATIGUE FROM THE ECONOMIC STANDPOINT. 299 



In 1893 Messrs. Mather & Piatt introduced the forty-eight hours 

 week at the Salford Ironworks. This was avowedly an experiment to 

 discover if hours could be shortened without endangering the trade, 

 and the result was that with the forty-eight hours week production was 

 actually increased relatively to the previous years when fifty-four or 

 fifty-three hours had been worked. 



Another experiment had been made recently at the Engis Chemical 

 "Works near Li^ge. The company had a sick benefit fund which was 

 constantly depleted, and the manager became alarmed. They tried the 

 effect of introducing three shifts oi eight hours. The output equalled 

 the previous output of ten hours' work, and the earnings, all piecework, 

 equalled the previous earnings. The increase of output and wages per 

 hour was about 33 per cent. The sick benefit accounts showed that 

 under the old system expenditure had exceeded receipts; while under 

 the new, receipts tended to exceed expenditure and that progressively. 

 The effects on health and sobriety were remarkable, and the cost of 

 production was reduced 33 per cent, per ton of roasted ore. 



Ernst Abb6, after becoming manager of the Carl Zeiss Works in 

 Jena, reduced the working day from mne to eight hours. In the result 

 it appeared the men earned by piecework on an average about 3 per 

 cent, more than they had earned in the previous year working nine 

 hours, and the earnmgs per hour increased in the ratio 100 : 116. The 

 men were unconscious of any special effort, and were surprised to find 

 their earnings increased. Abbe came to the conclusion that the in- 

 creased efficiency was physiological rather than psychological. If the 

 need of recuperation after exertion is neglected, the effect is like a daily 

 recurring deficit, and means actual loss in industry. 



Section III. — Do the Tests Vary with Fatigue ? — Are variations in 

 fatigue indicated by variations in the amount of output and 

 number of accidents ? 



That of the two tests we are mainly to use the output of work may 

 be expected to vary and to vary inversely with fatigue is suggested by 

 the very definition of fatigue as a diminution in the capacity for work. 

 That the occurrence of accidents may also be expected to vary and to 

 vary congruently with fatigue requires, however, some further proof. 



Accidents may be divided as to their causation into those in which 

 the human element entered and those in which it did not ; those whose 

 causes were partly psycho-physiological and those whose causes were 

 purely mechanical or otherwise ' external ' to the victim. 



In the causation of many accidents the psycho-physiological state 

 of the victim was probably one of the elements, though generally only 

 as a condition enabling some mechanical cause to take effect. A dan- 

 gerous psycho-physiological state may be either pennanent or tem- 

 porary, but in any case there is no reason why as a rule such a state 

 should be more usual later in the working period than earlier unless it 

 is connected with fatigue. 



Quite a few industrial accidents, however, are caused purely 

 mechanically by explosions, electric shocks, to which might be added 

 accidents due to animals ; others are attributable to recklessness, to the 



